- 标题
- 摘要
- 关键词
- 实验方案
- 产品
-
Solving Inverse Computational Imaging Problems using Deep Pixel-level Prior
摘要： Signal reconstruction is a challenging aspect of computational imaging as it often involves solving ill-posed inverse problems. Recently, deep feed-forward neural networks have led to state-of-the-art results in solving various inverse imaging problems. However, being task specific, these networks have to be learned for each inverse problem. On the other hand, a more flexible approach would be to learn a deep generative model once and then use it as a signal prior for solving various inverse problems. We show that among the various state of the art deep generative models, autoregressive models are especially suitable for our purpose for the following reasons. First, they explicitly model the pixel level dependencies and hence are capable of reconstructing low-level details such as texture patterns and edges better. Second, they provide an explicit expression for the image prior which can then be used for MAP based inference along with the forward model. Third, they can model long range dependencies in images which make them ideal for handling global multiplexing as encountered in various compressive imaging systems. We demonstrate the efficacy of our proposed approach in solving three computational imaging problems: Single Pixel Camera (SPC), LiSens and FlatCam. For both real and simulated cases, we obtain better reconstructions than the state-of-the-art methods in terms of perceptual and quantitative metrics.
关键词： lensless image reconstruction,MAP inference,Inverse problems,compressive image recovery,autoregressive models,deep generative models
更新于2025-09-10 09:29:36
-
[IEEE IGARSS 2018 - 2018 IEEE International Geoscience and Remote Sensing Symposium - Valencia, Spain (2018.7.22-2018.7.27)] IGARSS 2018 - 2018 IEEE International Geoscience and Remote Sensing Symposium - Monitoring Evapotranspiration with Remote Sensing Data and Ground Data Using Ensemble Model Averaging
摘要： Evapotranspiration (ET) can be mapped using thermal infrared and spectral reflectance data. Various ET models have been developed but there was no competitive evaluation of them over a large range of situations. Ensemble model averaging is a tool that can be used for deriving ET from multi-model simulations. In this study, we used bayesian model averaging, which consists in weighting each model according to their performances when deriving the ensemble average. It was applied to the monitoring of ET over a saltmarsh scrub area in South France from MODIS data. ET monitoring was improved (RMSE = 0.57 mm d-1) when using a weighted averaging procedure as compared to the performances of a simple average or to the performances of each individual model.
关键词： thermal infrared,energy balance,ensemble modelling,contextual models,evapotranspiration
更新于2025-09-10 09:29:36
-
Determination of hydrological roughness by means of close range remote sensing
摘要： The objective of the presented work was to develop a method to acquire Manning’s n by creating very high-resolution surface models with structure-from-motion methods. As hydraulic roughness is an essential parameter for physically based erosion models, a practical measuring technique is valuable during field work. Data acquisition took place during several field experiments in the Lainbach valley, southern Germany, and on agricultural sites in Saxony, eastern Germany, and in central Brazil. Rill and interrill conditions were simulated by flow experiments. In order to validate our findings stream velocity was measured with colour tracers. Grain sizes were derived by measuring distances from a best fit line to the reconstructed soil surface. Several diameters from D50 to D90 were tested with D90 showing best correlation between tracer experiments and photogrammetrically acquired data. Several roughness parameters were tested (standard deviation, random roughness, Garbrecht’s n and D90). Best agreement in between the grain size and the hydraulic roughness was achieved with a non-linear sigmoid function and D90 rather than with the Garbrecht equation or statistical parameters.
关键词： soil erosion models,hydrological roughness,close range remote sensing,Manning’s n,structure-from-motion
更新于2025-09-10 09:29:36
-
[IEEE 2018 25th IEEE International Conference on Image Processing (ICIP) - Athens, Greece (2018.10.7-2018.10.10)] 2018 25th IEEE International Conference on Image Processing (ICIP) - Image Aesthetics Assessment Using Composite Features from off-the-Shelf Deep Models
摘要： Deep convolutional neural networks have recently achieved great success on image aesthetics assessment task. In this paper, we propose an efficient method which takes the global, local and scene-aware information of images into consideration and exploits the composite features extracted from corresponding pretrained deep learning models to classify the derived features with support vector machine. Contrary to popular methods that require fine-tuning or training a new model from scratch, our training-free method directly takes the deep features generated by off-the-shelf models for image classification and scene recognition. Also, we analyzed the factors that could influence the performance from two aspects: the architecture of the deep neural network and the contribution of local and scene-aware information. It turns out that deep residual network could produce more aesthetics-aware image representation and composite features lead to the improvement of overall performance. Experiments on common large-scale aesthetics assessment benchmarks demonstrate that our method outperforms the state-of-the-art results in photo aesthetics assessment.
关键词： Feature Extraction,Pretrained Models,Image Aesthetics,Deep Learning
更新于2025-09-10 09:29:36
-
[IEEE 2018 International Symposium in Sensing and Instrumentation in IoT Era (ISSI) - Shanghai, China (2018.9.6-2018.9.7)] 2018 International Symposium in Sensing and Instrumentation in IoT Era (ISSI) - Image Denoising Using Asymmetric Gaussian Mixture Models
摘要： Finite mixture models are widely applied in image denoising because they have sound mathematical basis and the results are interpretable. In a manner of speaking, they can give a mathematical description of natural images by clustering. Usually assume that per-component of natural images follow a mixture of Gaussian(GMM) when doing image denoising. However, it is well- know that most of natural images are intricate and of which the distribution is highly non-Gaussian. So there remain problems that GMM cannot fix. In this paper, we introduce the asymmetric Gaussian mixture models into the finite mixture model, in which GMM is a special case. Asymmetric Gaussian Mixture (AGM) can model asymmetric distribution which is more conform to the data of natural images. We do some experiments in image denoising under different noise scales and types. The AGM can have better results compare to The GMM.
关键词： image denoising,priors,asymmetric Gaussian mixture models,expected patch log likelihood
更新于2025-09-10 09:29:36
-
[IEEE IGARSS 2018 - 2018 IEEE International Geoscience and Remote Sensing Symposium - Valencia, Spain (2018.7.22-2018.7.27)] IGARSS 2018 - 2018 IEEE International Geoscience and Remote Sensing Symposium - IBIS-ArcSAR: an Innovative Ground-Based SAR System for Slope Monitoring
摘要： In the last decade Ground-Based SAR (GB-SAR) technology has been demonstrated to be one of the most efficient tool for slope monitoring in open pit mines scenarios. However, especially for open pit mines application, one of the relevant limits is the azimuth coverage, being the mine often of elliptical shape. In this context, the IBIS-FM product of IDS GeoRadar company can reach about 80° of horizontal coverage, and can be overcome by IBIS-Rover product that can reach 270° of horizontal coverage but increasing the acquisition time. The recent IBIS-ArcSAR product developed by IDS GeoRadar can overcome these limits, offering a full 360° of horizontal coverage in less than 1 minute of acquisition time, with additional feature of a constant angular resolution. Moreover, by exploiting the new capability of MIMO (Multiple Input Multiple Output) Radar channels, the vertical height or Digital Elevation Model (DEM) of the mine can be reconstructed through simple and computationally fast interferometric processing algorithms, together with an enhanced focusing of 2D SAR Image that is an open issue in Arc-scanning SAR systems.
关键词： Interferometry,Ground-based Synthetic Aperture Radar,Digital Elevation Models,MIMO Radar,Radar Imaging
更新于2025-09-10 09:29:36
-
Comparison of geometry models in the study of perovskite heterostructures
摘要： Based on the first-principles calculations, the influence of different geometry models on the theoretical results about the electronic properties of LaAlO3/SrTiO3 perovskite heterostructures has been studied. The results show that the superlattice model can indicate n-type conductivity of interfaces, but it is limited to describe the insulator-metal transition of heterostructures. The sandwich and thin film models with the vacuum layer can take into account the surface termination effect on the electronic properties of the interface. The AlO2 or LaO atomic layer as the outmost surface termination is selected during constructing the model, leading to the distinct electronic properties of the heterostructures. The build-in electric filed appears in the AlO2-terminated structure, whereas it does not exist in LaO-terminated structure. Therefore, the AlO2-terminated structure can give a clear description about the insulator-metal transition of heterostructure and the charge transfer mechanism. The build-in electric filed, the atomic rumpling and surface termination effect of perovskite materials codetermine the conductivity of interfaces. The results would be helpful to construct theoretical model and understand the novel electronic properties during investigating other perovskite heterostructures.
关键词： perovskite heterostructure,surface termination,theoretical models,first-principles method
更新于2025-09-10 09:29:36
-
GPU-accelerated hierarchical Bayesian estimation of luminosity functions using flux-limited observations with photometric noise
摘要： We describe a C++ framework that uses graphical processing units (GPUs) to accelerate basic hierarchical Bayesian computation, and use it to compute the posterior distribution for the parameters of a galaxy luminosity function based on data with photometric noise and selection effects arising from a flux-limited detection threshold. In addition to estimating the population-level luminosity function parameters, the framework also provides estimates of the absolute magnitude of each object, which automatically correct for Eddington bias. To sample the posterior, we implement a Metropolis–Hastings-within-Gibbs algorithm that alternates between exploring the population-level and member-level parameters. The algorithm exploits conditional independence in hierarchical Bayesian models, which makes member-level exploration readily parallelizable on a GPU. The framework uses adaptive MCMC, automatically tuning Metropolis–Hastings proposal distributions on-the-fly. We present a simulation study demonstrating the accuracy and computational scaling of the algorithm. In addition, we compare hierarchical Bayesian estimation with maximum likelihood estimation (known to provide inconsistent estimates in this setting), providing a concrete demonstration of the benefits of using hierarchical models for luminosity function inference.
关键词： Parallel computing,Hierarchical Bayesian models,Graphical processing units (GPUs),Metropolis-within-Gibbs sampling,Markov processes,Astrostatistics
更新于2025-09-10 09:29:36
-
Unmanned Aerial Vehicle (UAV) photogrammetry produces accurate high-resolution orthophotos, point clouds and surface models for mapping wetlands
摘要： Unmanned Aerial Vehicle (UAV) photogrammetry has recently become a powerful tool that offers a viable alternative to traditional remote sensing systems, particularly for applications covering relatively small spatial extents. This paper presents results of a study that aimed at investigating the use of UAV photogrammetry as a tool for the mapping of wetlands. A multi-rotor UAV and a digital camera on a motion compensated gimbal mount were utilised for the survey. The survey of the 100ha study area at the Kameelzynkraal farm, Gauteng Province, South Africa took about 2? hours and the generation of the point cloud about 18 hours. Ground control points (GCPs) were positioned across the site to achieve geometrical precision and georeferencing accuracy. Structure from Motion (SfM) computer vision techniques were used to reconstruct the camera positions, terrain features and to derive ultra-high resolution point clouds, orthophotos and 3D models from the multi-view photos. The results of the geometric accuracy of the data based on the 20 GCPs were 0.018m for the overall and 0.0025m for the vertical root mean squared error (RMSE). The results exceeded our expectations and provided valuable, rapid and accurate mapping of wetlands that can be used for wetland studies and thereby support and enhance associated decision making to secure our future.
关键词： wetlands mapping,surface models,high-resolution orthophotos,point clouds,UAV photogrammetry,Structure from Motion
更新于2025-09-10 09:29:36
-
2018 Atmospheric Transmission Models-Modeling in Remote Sensing Meeting [Conference Reports]
摘要： The 2018 Atmospheric Transmission Models (ATM)–Modeling in Remote Sensing (MIRS) Meeting was held in Newton, Massachusetts, 4–6 June. This was a continuation of the long-running ATM meeting expanded by collaboration with the recently initiated IEEE Geoscience and Remote Sensing Society (GRSS) MIRS Technical Committee (TC). This report provides background on the meeting and the MIRS TC as well as summaries of the meeting presentations.
关键词： atmospheric retrievals,Modeling in Remote Sensing,radiative transfer,MODTRAN,molecular spectroscopy,Atmospheric Transmission Models
更新于2025-09-10 09:29:36